Examples with Schema org.apache.gobblin.source.extractor.schema.Schema used on opensource projects

Example 1 with Schema

use of org.apache.gobblin.source.extractor.schema.Schema in project incubator-gobblin by apache.

the class RestApiExtractor method extractMetadata.

@Override
public void extractMetadata(String schema, String entity, WorkUnit workUnit) throws SchemaException {
    log.info("Extract Metadata using Rest Api");
    JsonArray columnArray = new JsonArray();
    String inputQuery = workUnitState.getProp(ConfigurationKeys.SOURCE_QUERYBASED_QUERY);
    List<String> columnListInQuery = null;
    JsonArray array = null;
    if (!Strings.isNullOrEmpty(inputQuery)) {
        columnListInQuery = Utils.getColumnListFromQuery(inputQuery);
    }
    String excludedColumns = workUnitState.getProp(ConfigurationKeys.SOURCE_QUERYBASED_EXCLUDED_COLUMNS);
    List<String> columnListExcluded = ImmutableList.<String>of();
    if (Strings.isNullOrEmpty(inputQuery) && !Strings.isNullOrEmpty(excludedColumns)) {
        Splitter splitter = Splitter.on(",").omitEmptyStrings().trimResults();
        columnListExcluded = splitter.splitToList(excludedColumns.toLowerCase());
    }
    try {
        boolean success = this.connector.connect();
        if (!success) {
            throw new SchemaException("Failed to connect.");
        }
        log.debug("Connected successfully.");
        List<Command> cmds = this.getSchemaMetadata(schema, entity);
        CommandOutput<?, ?> response = this.connector.getResponse(cmds);
        array = this.getSchema(response);
        for (JsonElement columnElement : array) {
            Schema obj = GSON.fromJson(columnElement, Schema.class);
            String columnName = obj.getColumnName();
            obj.setWaterMark(this.isWatermarkColumn(workUnitState.getProp("extract.delta.fields"), columnName));
            if (this.isWatermarkColumn(workUnitState.getProp("extract.delta.fields"), columnName)) {
                obj.setNullable(false);
            } else if (this.getPrimarykeyIndex(workUnitState.getProp("extract.primary.key.fields"), columnName) == 0) {
                // set all columns as nullable except primary key and watermark columns
                obj.setNullable(true);
            }
            obj.setPrimaryKey(this.getPrimarykeyIndex(workUnitState.getProp("extract.primary.key.fields"), columnName));
            String jsonStr = GSON.toJson(obj);
            JsonObject jsonObject = GSON.fromJson(jsonStr, JsonObject.class).getAsJsonObject();
            // Else, consider only the columns mentioned in the column list
            if (inputQuery == null || columnListInQuery == null || (columnListInQuery.size() == 1 && columnListInQuery.get(0).equals("*")) || (columnListInQuery.size() >= 1 && this.isMetadataColumn(columnName, columnListInQuery))) {
                if (!columnListExcluded.contains(columnName.trim().toLowerCase())) {
                    this.columnList.add(columnName);
                    columnArray.add(jsonObject);
                }
            }
        }
        this.updatedQuery = buildDataQuery(inputQuery, entity);
        log.info("Schema:" + columnArray);
        this.setOutputSchema(columnArray);
    } catch (RuntimeException | RestApiConnectionException | RestApiProcessingException | IOException | SchemaException e) {
        throw new SchemaException("Failed to get schema using rest api; error - " + e.getMessage(), e);
    }
}

Example 2 with Schema

use of org.apache.gobblin.source.extractor.schema.Schema in project incubator-gobblin by apache.

the class JdbcExtractor method getDefaultWatermark.

/**
 * Schema of default watermark column-required if there are multiple watermarks
 *
 * @return column schema
 */
private JsonObject getDefaultWatermark() {
    Schema schema = new Schema();
    String dataType;
    String columnName = "derivedwatermarkcolumn";
    schema.setColumnName(columnName);
    WatermarkType wmType = WatermarkType.valueOf(this.workUnitState.getProp(ConfigurationKeys.SOURCE_QUERYBASED_WATERMARK_TYPE, "TIMESTAMP").toUpperCase());
    switch(wmType) {
        case TIMESTAMP:
            dataType = "timestamp";
            break;
        case DATE:
            dataType = "date";
            break;
        default:
            dataType = "int";
            break;
    }
    String elementDataType = "string";
    List<String> mapSymbols = null;
    JsonObject newDataType = this.convertDataType(columnName, dataType, elementDataType, mapSymbols);
    schema.setDataType(newDataType);
    schema.setWaterMark(true);
    schema.setPrimaryKey(0);
    schema.setLength(0);
    schema.setPrecision(0);
    schema.setScale(0);
    schema.setNullable(false);
    schema.setFormat(null);
    schema.setComment("Default watermark column");
    schema.setDefaultValue(null);
    schema.setUnique(false);
    String jsonStr = gson.toJson(schema);
    JsonObject obj = gson.fromJson(jsonStr, JsonObject.class).getAsJsonObject();
    return obj;
}

Example 3 with Schema

use of org.apache.gobblin.source.extractor.schema.Schema in project incubator-gobblin by apache.

the class JdbcExtractor method getSchema.

@Override
public JsonArray getSchema(CommandOutput<?, ?> response) throws SchemaException, IOException {
    this.log.debug("Extract schema from resultset");
    ResultSet resultset = null;
    Iterator<ResultSet> itr = (Iterator<ResultSet>) response.getResults().values().iterator();
    if (itr.hasNext()) {
        resultset = itr.next();
    } else {
        throw new SchemaException("Failed to get schema from database - Resultset has no records");
    }
    JsonArray fieldJsonArray = new JsonArray();
    try {
        while (resultset.next()) {
            Schema schema = new Schema();
            String columnName = resultset.getString(1);
            schema.setColumnName(columnName);
            String dataType = resultset.getString(2);
            String elementDataType = "string";
            List<String> mapSymbols = null;
            JsonObject newDataType = this.convertDataType(columnName, dataType, elementDataType, mapSymbols);
            schema.setDataType(newDataType);
            schema.setLength(resultset.getLong(3));
            schema.setPrecision(resultset.getInt(4));
            schema.setScale(resultset.getInt(5));
            schema.setNullable(resultset.getBoolean(6));
            schema.setFormat(resultset.getString(7));
            schema.setComment(resultset.getString(8));
            schema.setDefaultValue(null);
            schema.setUnique(false);
            String jsonStr = gson.toJson(schema);
            JsonObject obj = gson.fromJson(jsonStr, JsonObject.class).getAsJsonObject();
            fieldJsonArray.add(obj);
        }
    } catch (Exception e) {
        throw new SchemaException("Failed to get schema from database; error - " + e.getMessage(), e);
    }
    return fieldJsonArray;
}

Example 4 with Schema

use of org.apache.gobblin.source.extractor.schema.Schema in project incubator-gobblin by apache.

the class JdbcExtractor method extractMetadata.

@Override
public void extractMetadata(String schema, String entity, WorkUnit workUnit) throws SchemaException, IOException {
    this.log.info("Extract metadata using JDBC");
    String inputQuery = workUnitState.getProp(ConfigurationKeys.SOURCE_QUERYBASED_QUERY);
    if (hasJoinOperation(inputQuery)) {
        throw new RuntimeException("Query across multiple tables not supported");
    }
    String watermarkColumn = workUnitState.getProp(ConfigurationKeys.EXTRACT_DELTA_FIELDS_KEY);
    this.enableDelimitedIdentifier = workUnitState.getPropAsBoolean(ConfigurationKeys.ENABLE_DELIMITED_IDENTIFIER, ConfigurationKeys.DEFAULT_ENABLE_DELIMITED_IDENTIFIER);
    JsonObject defaultWatermark = this.getDefaultWatermark();
    String derivedWatermarkColumnName = defaultWatermark.get("columnName").getAsString();
    this.setSampleRecordCount(this.exractSampleRecordCountFromQuery(inputQuery));
    inputQuery = this.removeSampleClauseFromQuery(inputQuery);
    JsonArray targetSchema = new JsonArray();
    List<String> headerColumns = new ArrayList<>();
    try {
        List<Command> cmds = this.getSchemaMetadata(schema, entity);
        CommandOutput<?, ?> response = this.executePreparedSql(cmds);
        JsonArray array = this.getSchema(response);
        this.buildMetadataColumnMap(array);
        this.parseInputQuery(inputQuery);
        List<String> sourceColumns = this.getMetadataColumnList();
        for (ColumnAttributes colMap : this.columnAliasMap) {
            String alias = colMap.getAliasName();
            String columnName = colMap.getColumnName();
            String sourceColumnName = colMap.getSourceColumnName();
            if (this.isMetadataColumn(columnName, sourceColumns)) {
                String targetColumnName = this.getTargetColumnName(columnName, alias);
                Schema obj = this.getUpdatedSchemaObject(columnName, alias, targetColumnName);
                String jsonStr = gson.toJson(obj);
                JsonObject jsonObject = gson.fromJson(jsonStr, JsonObject.class).getAsJsonObject();
                targetSchema.add(jsonObject);
                headerColumns.add(targetColumnName);
                sourceColumnName = getLeftDelimitedIdentifier() + sourceColumnName + getRightDelimitedIdentifier();
                this.columnList.add(sourceColumnName);
            }
        }
        if (this.hasMultipleWatermarkColumns(watermarkColumn)) {
            derivedWatermarkColumnName = getLeftDelimitedIdentifier() + derivedWatermarkColumnName + getRightDelimitedIdentifier();
            this.columnList.add(derivedWatermarkColumnName);
            headerColumns.add(derivedWatermarkColumnName);
            targetSchema.add(defaultWatermark);
            this.workUnitState.setProp(ConfigurationKeys.EXTRACT_DELTA_FIELDS_KEY, derivedWatermarkColumnName);
        }
        String outputColProjection = Joiner.on(",").useForNull("null").join(this.columnList);
        outputColProjection = outputColProjection.replace(derivedWatermarkColumnName, Utils.getCoalesceColumnNames(watermarkColumn) + " AS " + derivedWatermarkColumnName);
        this.setOutputColumnProjection(outputColProjection);
        String extractQuery = this.getExtractQuery(schema, entity, inputQuery);
        this.setHeaderRecord(headerColumns);
        this.setOutputSchema(targetSchema);
        this.setExtractSql(extractQuery);
        // this.workUnit.getProp(ConfigurationKeys.EXTRACT_TABLE_NAME_KEY,
        // this.escapeCharsInColumnName(this.workUnit.getProp(ConfigurationKeys.SOURCE_ENTITY),
        // ConfigurationKeys.ESCAPE_CHARS_IN_COLUMN_NAME, "_"));
        this.log.info("Schema:" + targetSchema);
        this.log.info("Extract query: " + this.getExtractSql());
    } catch (RuntimeException | IOException | SchemaException e) {
        throw new SchemaException("Failed to get metadata using JDBC; error - " + e.getMessage(), e);
    }
}

Example 5 with Schema

use of org.apache.gobblin.source.extractor.schema.Schema in project incubator-gobblin by apache.

the class JdbcExtractor method buildMetadataColumnMap.

/**
 * Build metadata column map with column name and column schema object.
 * Build metadata column list with list columns in metadata
 *
 * @param array Schema of all columns
 */
private void buildMetadataColumnMap(JsonArray array) {
    if (array != null) {
        for (JsonElement columnElement : array) {
            Schema schemaObj = gson.fromJson(columnElement, Schema.class);
            String columnName = schemaObj.getColumnName();
            this.metadataColumnMap.put(columnName.toLowerCase(), schemaObj);
            this.metadataColumnList.add(columnName.toLowerCase());
        }
    }
}